Generator wind temperature bearing vibration

How accurate are vibration signals in wind turbine generators?

Evaluation of the bearing in horizontal, axial, and vertical, with an accuracy of 91%. Non-stationary vibration signal in the absence of an external hardware sensor. Acoustic and vibration signals for different fault cases for gearbox in Wind turbine generators, with hybrid ensemble is developed by stacking the RF, DT, KNN.

How to detect bearing faults in wind generators?

Early detection of bearing faults in wind generators, it Utilized a multi-stage approach involving Random Forest, XGBoost, Light XGB, and Logistic Regression, followed by probability scores and optimal features with a search grid validation; as ensembled method. Torsional sensors are not common in the drivetrain.

Are wind turbine bearing vibration condition monitoring studies relevant?

In the literature review into the wind turbine bearing condition monitoring field by de Azevedo, three out of 38 relevant bearing vibration condition monitoring studies were performed using field data compared with test rig data.

Can a vibration sensor detect a wind turbine fault?

Currently, the state of the art in wind turbine fault detection is limited to vibration as the sole variable. However, vibration sensors can only detect 5–20% of torsional vibration in the drivetrain, caused by the dynamic and natural frequency of the system .

Can a WPT measure a wind turbine bearing failure?

Mollasalehi et al implemented the WPT to find and separate the high-frequency content in vibration measurements from a wind turbine generator bearing failure and demodulated these frequency bands and applied the FFT to correlate to the bearing fault frequencies. However, no direct comparison to only using the FFT was performed.

How do we classify a wind turbine generator?

During feature extraction, vibration comparison was used to evaluate the behavior of the wind turbine generator (WTG) and classify it into three categories: 0 for normal operation, 1 for gearbox failure, and 2 for bearing failure. The evolutionary information consisted of a matrix of “n" events with three classifications.